Monday, August 27, 2007

Lumpers and Splitters

Thank you for coming to my presentation today. The title of my talk is Lumpers and Splitters: A NeoAgrarian Syndicalist Approach to Procrustean Obfuscatory Hand-Waving-entarianism

Slide 1: Outline.
I will be covering this blog post in three steps.
Step 1: Statement of the Problem
Step 2: A detour into linguistica historica
Step 3: Discussion, Limitations of the Study, and Future Research

Statement of the Problem.
Are you a lumper or a splitter?

Literature Review:
J (2007: comment 2) quotes Anonymous, citing the amusing quotidian: "There are two kinds of people in the world: those who divide the world up into two kinds of people and those who don't."

Research Hypothesis:
I am a lumper.

Methodology:
To explore this question, I asked myself silently if I was a lumper or a splitter.

Results:
I said I was a Lumper.
Reliability of this result is poor, but then it's all the same to me.

Discussion: (the real post begins)
J's comment made me think of historical linguistics, which is the subfield that studies how languages developed through time. Often the goal is to create the great genetic tree - a tree which shows which languages evolved from which earlier languages. So if you've ever seen a tree which has Italic, Greek, Sanskrit, Celtic, Germanic, etc., descending from an Indo-European language, that tree is the fruit of years and years of labor of a bunch of linguists.

So how is a tree built? The primary methodology for 150 - 200 years is comparing words. Basically, if they seem similar, then the languages might be related. However, there are lots of things to watch out for.

1) Some words are similar by chance. I believe that "ahi" is the word for tuna in both Japanese and Hawaiian. But that's just by chance. There's always a few words here and there.

2) Some words come out similar for reasons having nothing to do with history. Mama is mother in English and Chinese (and in many other languages). However, this most likely is because the words "mama", "papa", and "dada" are the easiest words for a baby to say physically. Just open and close your mouth while making sound and it comes out kind of like "baba." Apparently, cultures have a tendency to take that and make it a parental name. (Apparently, "mama" is father in Georgian (the country).) Some other words also have a reason for their sound. In English, we say "cat", but in Chinese a cat is a "mao" (mow). I've never looked it up, but Germanic "wolf" sure sounds like woof woof to me. Words like these then are not very useful in tracing history, because two independent cultures might easily come up with woof for dogs.

3) Here's the big one that drives linguists insane. Borrowing. Unrelated languages can borrow words due to cultural contact. Sometimes it's obvious like with tsunami, kahuna, or quid pro quo, but most often not. The word "wine" was borrowed from Latin a couple thousand years ago. It is not evidence, however, that English is historically descended from Latin. English is not a new form of Latin. Similarly, English has words like "cardiac", "cardiologist", and "cardiovascular," which are all related to French "coeur" for heart. But again, this is not evidence that English is a historical descendant of French. Instead, they are more like cousins in Alabama; i.e., they are related through their grandparents but just can't stay out of each other's bed. "Card" was borrowed from French, not inherited. The way you rule out such borrowings, when trying to trace history, is by looking not just at what's common, but what changes were in common. In Germanic, inherited "p"s changed systematically across many words into "f"s. "D"s changed into "th"s. And so a "pader" in Indo-European became a "father" in Germanic. (This is Grimm's Law from the mid-19th c.) Similarly "k" became "h", so card became heart. The fact that the c/k sound is still there in "cardiac" tells you that it is a recent borrowing (like 1066) and not an inheritance. Sound change is supposed to be regular (tough it's not).

Finally coming back to the beginning. In historical ling, when people find some commonalities, but it's not overwhelming, you end up with natural lumpers and splitters. One commonality for a lumper is evidence to add a line to the tree. To a splitter, only finding one commonality is a reason to cross the line out. Due to this, you can find some very different sorts of trees. In one famous tree, a linguist attempted to argue that there were only three families of American languages from Alaska to Argentina. Others will give you 50.

I don't do historical ling, but I find myself a peculiar lumper in the work on politeness and Korean apologies that I'm doing. People have come up with generalizations about Western versus East Asian cultures, wherein East Asian ones are supposed to be "collectivist" and "shame societies" while Western ones are "individualist" and "guilt societies". While these are kind of useful, I'm always trying to show how East Asian culture has more individualist tendencies than it's given credit for, and Western ones are more collectivist than usually recognized. It's kind of a peculiar lumping technique. I first apparently try to split by showing diversity in the society, and then since we now see the splits, we can do the lumping and see how humans are similar across cultures. Lumpo ergo splitto. I lump, therefore I split.

(There was supposed to be a Monty Python and Evil Editor discussion here, but this is far too long and its back to work. SPLITTER!!)

1 comment:

Ello said...

And now for something completely different...

I don't know if I understood any of this but I think I would rather be a lumper than a splitter. Seems easier. So now, please post the Monty Python post! I LOVE MONTY!